~10% hallucinated sentences: guardrails for lifelogging with LLMs

thejaan · x · 2026-09-24

In a follow-up to his lifelogging thread, Jaan Altosaar details that about 10% of the LLM-generated sentences are made up (e.g. one claimed his dad was recovering from surgery in Estonia), so he listens with vigilance. His guardrails: no memory across sessions, no follow-up questions to avoid reassurance-seeking, a fresh session every 5-10 weeks, and treating it as no substitute for therapy.

Related event: Jaan Altosaar Turns Six Months of Journals, Sleep and Finance Data into AI-Generated Podcasts with Claude(5 posts)→

Original post →

More from Apps

Apps channel →